On Tue, 23 Aug 2005 23:03:05 +0200, Andreas Barth <[EMAIL PROTECTED]> said:
> * Russ Allbery ([EMAIL PROTECTED]) [050823 22:58]:
>> Goswin von Brederlow <[EMAIL PROTECTED]> writes:
>> > It is out of date since it does not explain ~ yet. Maybe, if you
>> > have the time and since you just looked a
Andreas Barth <[EMAIL PROTECTED]> writes:
> * Russ Allbery ([EMAIL PROTECTED]) [050823 22:58]:
>> Goswin von Brederlow <[EMAIL PROTECTED]> writes:
>> > It is out of date since it does not explain ~ yet. Maybe, if you have
>> > the time and since you just looked at the matter closely anyway, you
>>
also sprach Russ Allbery <[EMAIL PROTECTED]> [2005.08.23.2257 +0200]:
> I'm certainly willing to do so, but I thought that policy wasn't ready to
> change yet. Wasn't it waiting on implementation of that feature in dak,
> which is currently using ~ internally for something else?
Yes, APT and dpkg
* Russ Allbery ([EMAIL PROTECTED]) [050823 22:58]:
> Goswin von Brederlow <[EMAIL PROTECTED]> writes:
> > It is out of date since it does not explain ~ yet. Maybe, if you have
> > the time and since you just looked at the matter closely anyway, you
> > could draw up a few lines and send a patch?
>
Goswin von Brederlow <[EMAIL PROTECTED]> writes:
> It is out of date since it does not explain ~ yet. Maybe, if you have
> the time and since you just looked at the matter closely anyway, you
> could draw up a few lines and send a patch?
I'm certainly willing to do so, but I thought that policy w
Russ Allbery <[EMAIL PROTECTED]> writes:
> Thanks for the confirmation!
>
> martin f krafft <[EMAIL PROTECTED]> writes:
>> also sprach Russ Allbery <[EMAIL PROTECTED]> [2005.08.23.1908 +0200]:
>
>>> Is there a document anywhere outside of the dpkg source that explains
>>> the algorithm for how ver
Thanks for the confirmation!
martin f krafft <[EMAIL PROTECTED]> writes:
> also sprach Russ Allbery <[EMAIL PROTECTED]> [2005.08.23.1908 +0200]:
>> Is there a document anywhere outside of the dpkg source that explains
>> the algorithm for how version numbers are ordered by the archive
>> software
dpkg --compare-versions provides exactly the ordering that I want, namely
that 1.4rc1 < 1.4.0 so by omitting the final patch number in the RC
revision I can use the correct upstream version without using epochs or
strange-looking version numbers. However, since this is a bit of an edge
case
also sprach Russ Allbery <[EMAIL PROTECTED]> [2005.08.23.1908 +0200]:
> case, I wanted to double-check and be sure that dpkg --compare-versions is
> the canonical ordering for version numbers. I'm pretty sure it is, but
> better safe than sorry to check.
Yes.
> Is th
I need to run some commands in the postinst of ifmail only if the user
is upgrading a previous. I tried:
if [ "$1" == "upgrade" -a -n "$2" -a "$2" != "" ]; then
if dpkg --compare-versions $2 le 2.12tx8.6; then
[...]
fi
fi
but it does not wo
mpatibility with control file syntax).
BTW, that's a typo 'tread' should be 'treat', I think
Yet:
$ dpkg --compare-versions 1.2.3 '<<' 4.5.6
dpkg: --cmpversions bad relation
Type dpkg --help for help about installing and deinstalling packages [*];
Use dse
I have been trying to use dpkg --compare-versions in a perl script I am
working on. The command seems to execute ok, but I get nothing back from
it. How do I tell if the command has been successful or not?
TIA,
Dwarf
--
aka
12 matches
Mail list logo
